Interactive E-book

Initial situation

Traditionally, the courses are presented in the form of classic ex cathedra courses with a visual support (slides) containing a summary of theoretical elements and illustrations (examples). The teacher wished to modify this overly static course format and make it more dynamic in order to make it more consistent with what is actually done in the world of research, i.e., learning and producing knowledge simultaneously. In the field of statistics, the fundamental notions are common to all data analysis methods, and they can be stated in a very general way. On the other hand, the implementation of these notions in the practice of data analysis (inferential statistics) is very diverse and can be applied in as many ways as there are practical problems. The teacher wished to focus on learning statistical methodology (theoretical part) towards data-driven decision making and to lead the student towards the greatest possible autonomy.

Project implementation

An HTML document was produced using RStudio software containing all the course material (text, R code, references and links, practical analyses, simulation exercises, etc.) in the form of an interactive e-book. This e-book was then made available to students on GitHub (a web service for hosting and managing software development) providing them with a common work environment dedicated to the course topic. This space, in addition to gathering the course resources, allows students to collaborate and exchange more easily among themselves and to discover and appropriate the use of the services offered by the platform widely used by data scientists (e.g., forum, R package development).

The weekly courses are divided into two parts, a course dedicated to the learning of concepts (more theoretical part) and a seminar dedicated to the supervision of the practical application of data analysis. In concrete terms, students come to class having already read the chapter of the week. This chapter is designed to expose them to the theoretical concepts, from the fundamental notions to the different methods of data analysis and their limits of application. Face-to-face time is used to review and expand on the material studied in advance, and to answer students’ questions. The same approach is used in the seminar part where students validate their practical application of data analysis on concrete datasets, without disciplinary limits, using the open source statistical software R.

In order to reinforce the link between the learning of methodology and its application, in parallel to the classroom sessions, students, in groups of 2 to 3, carry out data analysis projects during the semester. The aim of this approach is to expose and train students to produce knowledge based on data, from the research question(s) that motivated the collection of data, to the choice of analysis methods, with a critical mind on the limits of the conclusions/decisions that can be drawn from the data. In concrete terms, they have a set of data and will carry out a complete analysis by testing different methods in order to provide one or more interpretations of the observed data and compare the methods used. They must also be able to argue their choice of analysis methods. In recognition of their effort, the best analyses are included in the e-book with acknowledgement to their authors.

The evaluation of the course is made up of a written report on the group project (focusing mainly on the reflective approach that led to the analysis presented) and an oral evaluation of the project and some of the fundamental notions seen in class.

Thoughts and advice

GitHub is an extremely practical tool offering many possibilities of use: almost automatic creation of a web page, discussion forum, possibility of working with several people on a document or on the development of an R package, etc. It also allows the teacher to have a look on what has been done and by which student, which allows to give personalized feedback, but also to know if their contributions are equal in the framework. It also allows the teacher to have a look on what has been done and by which student, which allows to give personalized feedbacks, but also to know if their contributions are equal in the group work.

The e-book format for the course material reduces teacher/student dependency in the sense that the material is self-sufficient: the student is more autonomous, but has to adopt a more responsible (and autonomous) approach to learning, and the teacher becomes the support (and not the driver) of the student’s learning. In addition, and as an additional motivation, the e-book format allows for the participative construction of knowledge since some of the students’ productions can be included in the e-book for subsequent courses, without any limit in quantity. The course material is thus made dynamic and very portable, which would allow it to be easily adapted to be given entirely at a distance, or even shared with teachers in other institutions.

Student feedback

The pedagogical approach is still too unusual, which leads students to believe that the course requires a lot of preparation (beforehand) compared to other courses. They think that learning is more effective when the teacher repeats the open book content in the classroom.